Progression of biological markers in spinocerebellar ataxia type 3: longitudinal analysis of prospective data from the ESMI cohort

Abstract

Background: Spinocerebellar ataxia type 3 (SCA3) is an autosomal dominantly inherited adult-onset disease. We aimed to describe longitudinal changes in clinical and biological findings and to identify predictors for clinical progression. Methods: We used data from participants enrolled in the ESMI cohort collected between Nov 09, 2016 and July 18, 2023. The data freeze included data from 14 sites in five European countries and the United States. We assessed ataxia with the Scale for the Assessment and Rating of Ataxia (SARA). We measured disease-specific mutant ataxin-3 protein (ATXN3) and neurofilament light chain (NfL) in plasma and performed MRIs. Data were analysed by regression modelling on a timescale defined by onset. The onset of abnormality of a marker was defined as the time at which its value, as determined by modelling, exceeded the mean ± 2 SD of healthy controls. To study responsiveness of markers, we determined the sensitivity to change ratios (SCSs). Findings: Data from 291 SCA3 mutation carriers before and after clinical onset and 121 healthy controls were included. NfL levels became abnormal in SCA3 mutation carriers more than 20 years (-21.5 years [95% CI n.d.-9.5]) before onset. The earliest MRI abnormality was volume loss of medulla oblongata (-4.7 years [95% CI n.d.-3.3]). The responsiveness of markers depended on the disease stage. Across all stages, pons volume had the highest responsiveness with an SCS of 1.35 [95% CI 1.11-1.78] exceeding that of SARA (0.99 [95% CI 0.88-1.11]). In SCA3, lower age (p = 0.0459 [95% CI of slope change -0.0018 to 0.0000]) and lower medulla oblongata volume (p < 0.0001 [95% CI of slope change -0.0298 to -0.0115]) were predictors of SARA progression. Interpretation: Our study provides quantitative information on the progression of biological markers in SCA3 mutation carriers before and after onset of ataxia, and allowed the identification of predictors for clinical progression. Our data could prove useful for the design of future clinical trials.